Audio Captioning on AudioCaps Challenging
0.0541BLEU-4Baseline
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| BaselineSystem=Baseline2025.09 | 0.0541 | 0.3357 | 0.4565 | 0.1441 | 0.2546 | 0.5482 | 40.86 | 46.07 | |
| RLHF-based audio captioning frameworkSystem=+RLHF2025.09 | 0.0498 | 0.3465 | 0.5007 | 0.3571 | 0.2572 | 0.6025 | 59.14 | 53.93 |